CN108811176A - A kind of centralized contention resolution of wireless internet of things random multi-access channel - Google Patents

Abstract

A centralized conflict resolution method for random multiple access of the wireless Internet of Things. All stations with data to be sent randomly select a time slot after receiving the beacon frame sent by the gateway and send a request access signal to the gateway. The gateway coordinates In transmission order, stations choose to send data or contend for the channel again. Time-division multiplexing is used to divide the shared channel occupancy period into four periods: downlink announcement, uplink contention, downlink confirmation and uplink data transmission. During the downlink notification, the gateway broadcasts the beacon and allocates the number of contention time slots during the uplink contention. The access station randomly selects the contention time slots during the uplink contention and sends a request access signal to the gateway. The gateway broadcasts all contention time slots during the downlink confirmation. The feedback information of the contention status, the access station chooses to send data immediately, or delay sending data, or contend for the channel again, or delay to contend for the channel again according to the contention status. This application ensures that all stations send data successively without interfering with each other, thereby improving the throughput performance of the network.

Description

A Centralized Conflict Resolution Method for Random Multiple Access in Wireless Internet of Things

technical field

The present application relates to the field of wireless communication, in particular to a centralized conflict resolution method for random multiple access in the wireless internet of things.

Background technique

In a wireless communication system, data is transmitted through a wireless medium as a transmission channel. In IoT applications, a large number of nodes often access the channel at the same time. At this time, data transmission may collide and conflict, resulting in a decrease in channel utilization and throughput performance of the communication system.

The Media Access Control (MAC) protocol of wireless multiple access is responsible for the allocation of shared wireless channels and the scheduling of communication resources occupied by nodes, ensuring that nodes in the network can share communication resources reasonably and efficiently. At present, the industry proposes a Frame Slotted ALOHA (Frame Slotted ALOHA, FSA) scheme that does not require coordination to solve the conflict caused by channel access in the RFID system. In order to alleviate the problem of tag collision, FSA estimates the number of tags generated by collision to dynamically adjust the number of time slots in each frame, and builds a query tree based on the query of tag subgroups to prevent conflicts.

Sigfox uses the random frequency-time multiplexing method to propose a channel collision judgment method, which allows each node to select a frequency in the continuous available frequency band for asynchronous transmission. The channel collision judgment method does not need channel detection, network time synchronization and high-precision oscillator, but is susceptible to interference and collision.

The channel access method of the terminal in the LoRa system adopts the pure ALOHA working mechanism, and the terminal directly sends without channel detection. When the load increases, the probability of channel collision increases, and the throughput performance of the system may drop sharply.

Contents of the invention

This application proposes a centralized conflict resolution method for wireless Internet of Things random multiple access, the specific method is divided into the following steps:

(1) The gateway (Gate-Way, GW) broadcasts the beacon frame (BeaCoN, BCN) to all end stations (End-Device, ED) to complete time synchronization, followed by a fixed-length competition window (Collision Window, CW);

(2) After receiving the beacon, the station that has data to be sent and is in the ready state sends a signal requesting access to the gateway, that is, the time slot segment occupied by the requesting access;

(3) The gateway broadcasts the occupancy of each time slot after the CW is closed;

(4) After receiving the feedback frame (Feed-Back, FB) from the gateway, the site that has sent the transmission request judges the occupancy of the time slot: if it is free, it enters the data transmission queue (Data Transmission Queue, DTQ) and waits to send Data frame, otherwise enter the conflict resolution queue (Contention Resolution Queue, CRQ) and wait for the next CW to open to compete for the time slot again;

(5) The station in DTQ estimates the backoff time according to its position in the queue, enters the sleep state within this time, and sends data as soon as the time is up; each data transmission is completed, followed by CW and feedback frames of the same length , the stations in the CRQ compete for the time slot again when the next CW arrives;

The multi-site channel access control scheme uses time division multiplexing to divide the shared channel into four periods: downlink announcement, uplink contention, downlink confirmation and uplink data transmission;

During the downlink notification period, the central gateway broadcasts the beacon and allocates the number of contention slots during the uplink contention period;

The access station randomly selects a contention time slot during the uplink contention period and sends a request access signal to the gateway;

During the downlink confirmation period, the gateway broadcasts the contention status of all contention slots;

During the uplink data transmission period, the access station chooses to either send data immediately, or delay sending data, or contend for the channel again, or delay to contend for the channel again, according to the contention state.

Further, the number of contention time slots in the downlink announcement period is determined by the gateway according to the number of access stations and the number of uplink data requests, and the value ranges from 3 to 32.

Further, for the random selection of the contention time slots in the uplink contention period, the access station independently executes the uniform distribution pseudo-random number calculation, and the value range is distributed between 1 and the maximum number of time slots.

Further, the contention status of the downlink acknowledgment period is determined by the gateway according to the received access request signal from the access station, and is one of success, conflict and idle.

Further, during the uplink data transmission period, if the access station chooses to send data, it enters DTQ; the condition for sending data immediately is that the contention slot state selected by the station is successful and is at the head of the DTQ queue; the condition for delaying sending data is, The contention slot status selected by the station is successful but not at the head of the DTQ queue.

Further, during the uplink data transmission period, if the access station chooses to contend for the channel, it enters CRQ; the condition for competing for the channel again is that the contention time slot selected by the station is in conflict and is at the head of the CRQ queue; The condition is that the contention slot state selected by the station is conflicted but not at the head of the CRQ queue.

The communication control process described in the method of the present invention determines whether a conflict occurs by detecting the occupancy of the time slot, and retransmits the stations requesting access in groups, so as to ensure that all stations send data successively in a non-interfering manner, which can improve the throughput performance of the network .

Description of drawings

FIG. 1 is a schematic diagram of a multi-site channel access scenario in this application.

Fig. 2 is a communication process diagram of the successful multi-site access proposed by the present application.

Fig. 3 is a diagram of a communication process in a multi-site access conflict scenario proposed by the present application.

Fig. 4 is a flow chart of the gateway operation for multi-site access proposed in this application.

FIG. 5 is a flow chart of site operations for multi-site access proposed in this application.

FIG. 6 is an example of random selection of time slot numbers for 5 stations contending for a shared channel given in this application.

Fig. 7 is a timing example of 5 stations contending for a shared channel given by the present invention.

Detailed ways

The technical solution of the present invention will be further described in detail below in conjunction with the accompanying drawings.

Fig. 1 is a communication scenario between multi-sites and gateways in the present invention. 102 represents a gateway (GW), and 104, 106, and 108 represent end stations (EDs). Usually, in a network, multiple EDs send access request signals to the GW at the same time.

Fig. 2 is a communication process diagram of multi-site successful access proposed by the present invention. Assume that only one ED has data to send in that BCN cycle. It should be noted that a station that has data to send in each BCN cycle must handle all conflicts and complete sending all data before the next BCN arrives. Therefore, CRQ and DTQ are empty after each BCN cycle ends. According to the solution proposed by the present invention, it is divided into four communication stages.

Phase 1: The GW broadcasts the BCN frame to the downlink channel, which corresponds to step 206 in the figure.

Phase 2: GW opens the competition window CW (here it is assumed that CW consists of 3 time slots, and the cycle has been set), corresponding to 208 in the figure; ED that is ready to send data switches to the receiving state, and completes clock synchronization after receiving BCN , and at the same time randomly select a time slot (assuming that the first time slot is selected) and send a signal requesting access to the GW, corresponding to 210 in the figure.

Phase 3: As shown in 212 in the figure, after receiving all the requests from the ED, the GW waits until the CW is closed, and at this time, the GW broadcasts the occupancy of the time slot to the downlink channel.

Stage 4: ED receives the feedback information sent by GW and determines that time slot 1 is exclusively occupied by it, so ED enters DTQ; because only one station has a signal requesting access during this time period, and there is data to be sent in the previous time period The station has finished sending in the last BCN cycle, so ED is at the head of the DTQ queue and can send data without waiting, corresponding to 214 in the figure.

Fig. 3 is a diagram of the communication process in the multi-site access conflict scenario proposed by the present invention. Suppose there are two EDs (ED1, ED2) ready to send data in this BCN cycle. According to the multi-site conflict resolution method proposed by the present invention, it is divided into the following nine communication stages.

Phase 1: The GW broadcasts the BCN frame to the downlink channel, corresponding to 308 in the figure.

Phase 2: As shown at 310 in Figure 3, GW opens the contention window CW (CW consists of three fixed-period contention time slots); ED1 and ED2 randomly select a time slot after receiving the BCN (assuming time slots are selected at the same time). slot 2) and send a signal requesting access to the GW, corresponding to steps 312 and 314 respectively.

Phase 3: After the GW receives all requests from the ED during the CW opening period, it broadcasts the contention of the time slot through the downlink channel after the CW is closed, corresponding to 316 in the figure.

Phase 4: After receiving the feedback information from the GW, ED1 and ED2 find that the selected time slot 2 is contended by multiple stations, that is, a conflict occurs, and they enter CRQ waiting. At this time, since only ED1 and ED2 have data to send in the network, and no other ED has a signal to request access, that is, DTQ is empty, so GW turns on CW regularly after reaching the maximum back-off time, corresponding to 318 in the figure . ED1 and ED2 are at the head of the CRQ queue, and they compete for the time slot again after waiting for the same length of time (assuming that ED1 selects time slot 3 this time, and ED2 selects time slot 2), corresponding to 320 and 322 in the figure.

Phase 5: After the CW ends, the GW broadcasts the time slot contention situation again, which corresponds to 324 in the figure.

Phase 6: Since no other stations need to compete for time slots in this BCN cycle, all stations in the DTQ need to successfully send data to the GW. Because ED2 selects time slot 2, and ED1 selects time slot 1, according to the principle that the station with the first selected time slot enters DTQ first, so in DTQ, ED2 is one bit ahead of ED1, that is, ED2 sends data first, and ED1 sends data later data. At this time, there are only ED1 and ED2 in DTQ, so ED2 is sent directly without waiting (corresponding to 326 in the figure).

Stage 7: As shown in 328, after the GW receives the data, it still needs to open the CW. Because the conflict between ED1 and ED2 has been resolved in this BCN cycle, there is no ED to compete for the time slot.

Phase 8: After the CW ends, the GW broadcasts the FB on the downlink channel.

Stage 9: At this time, ED1 is at the head of the DTQ team and does not need to back off. It sends data directly after receiving the feedback message, corresponding to 332 in the figure. In this way, in this BCN cycle, all data can reach the GW successfully, and the conflict is resolved.

Fig. 4 is a flow chart of gateway operation for multi-site access proposed by the present invention. Step 402 represents the start of operation. The gateway first needs to set the cycle of BCN and CW. Then the broadcast BCN completes the clock synchronization, and at the same time opens the CW, as shown in step 404 . During the opening of the CW, the stations can compete for time slots. Step 406 indicates that the gateway determines whether the CW needs to be closed. If not, then execute step 408. Otherwise, jump to step 410. The gateway sends the occupancy of the FB broadcast time slot (subsequent operations are determined by site complete). During this process, the gateway needs to judge whether the BCN period is 0, as shown in step 412 . If yes, go to step 404 to send the next frame BCN, otherwise go to step 414 . Step 414 represents reducing the period of the BCN while resetting the period of the CW.

Fig. 5 is a flow chart of site operation for multi-site access proposed by the present invention. Step 502 indicates that all stations receive the BCN frame sent by the gateway for synchronization. Then all stations judge whether there is data to send (step 504), and if so, step 506 is executed sequentially, that is, a signal for requesting access is sent. Otherwise, jump to step 534 and do not perform any operation. After receiving the FB sent by the gateway (step 508), the station executes step 510 to judge whether there is a conflict in the selected time slot. If the time slot is selected by multiple stations, then jump to step 512, and all stations that select the time slot enter the same section of CRQ together. Each station in the CRQ will execute step 514 to determine whether the station is at the head of the CRQ team. If yes, execute step 516 to determine whether the next CW is enabled. If it is enabled, jump to step 506, and perform time slot contention again during the opening period of the CW. Otherwise, execute step 518, sleep and wait. If the selected time slot is exclusively occupied by the station, step 520 is executed, and the station enters DTQ. Each station of the DTQ will execute step 522 to judge whether the station is at the head of the DTQ team. If yes, continue to judge whether the station has received FB at this time, as shown in step 526 . If yes, execute step 530 to send data. While sending data, each station judges whether the current data has been sent, as shown in step 532 . If the transmission is completed, the communication at this stage can be judged to be over (step 534).

Consider the situation that 5 ED stations (ED1~ED5) compete for 3 time slots at the same time. The slot numbers randomly selected by each station are listed in Figure 6.

According to the selection result, there is no conflict in slot 2. Therefore, after receiving the FB from GW, ED2 sends data immediately. After the GW receives the data, it opens the CW. At this time, ED1 and ED4 contend for the time slot again. After the CW is closed, the GW broadcasts the time slot contention situation. After receiving the feedback information, ED1 and ED4 find that there is no conflict in the selected time slot, and enter DTQ to wait for sending data. Since DTQ was empty before, ED4 is at the head of the team and sends data immediately. After the GW receives the data, it opens the CW again, and broadcasts the FB frame to the downstream channel when the CW ends. After ED3 and ED5 receive the feedback information, they judge to enter DTQ and wait to send data. At this point, there are 3 stations in DTQ, namely ED1, ED5 and ED3. Since ED1 is at the head of the queue, data can be sent immediately. Likewise, GW opens CW as usual and broadcasts FB after CW closes. After receiving the FB, the ED5 at the head of the team sends data to the GW. After receiving the data, the GW opens the CW and broadcasts the slot contention after the CW is closed. ED3 can send data after receiving the feedback information. In this way, all five stations have completed the transmission of uplink data. The sequence flow of processing is shown in Fig. 7 .

It can be seen from Figure 7 that the total time for the five ED stations to occupy the shared channel is:

t = t _BCN + ( t _cw + t _FB + t _DATA )×5

Wherein, t _BCN is the occupied time of BCN transmission, t _CW is the CW cycle, t _FB is the occupied time of FB transmission, and t _DATA is the data transmission time.

Suppose the BCN length is 20B, the CW is composed of three time slots of 66.7μs, the FB length is 12B, and the transmitted data packet size is 10B. For the data transmission rate of 2000 b/s, and ED generates a data packet every 10 s, then:

tBCN ₌ 20×8/2000 = 0.08 (s)

t _FB = 12×8/2000 = 0.048 (s)

t _DATA = 10×8/2000 = 0.04 (s)

so,

t = 0.08+(66.7×3×10 ^-6 +0.048+0.04)×5 = 0.521 (s)

The system normalized load is G = 5 × 0.521/10 = 0.2605. Regardless of transmission errors, the system throughput of the present invention is:

S = G = 0.2605

Compared to pure ALOHA access control, the throughput is:

S = G × exp(-2 G ) = 0.1547

It can be seen that the method of the present invention has a performance improvement of 68.4% compared with pure ALOHA under the same load.

The above descriptions are only preferred embodiments of the present invention, and the scope of protection of the present invention is not limited to the above embodiments, but all equivalent modifications or changes made by those of ordinary skill in the art according to the disclosure of the present invention should be included within the scope of protection described in the claims.

Claims (6)

1. A centralized conflict resolution method for wireless internet of things random multiple access, characterized in that: the specific method is divided into the following steps: (1) The gateway (Gate-Way, GW) broadcasts the beacon frame (BeaCoN, BCN) to all end stations (End-Device, ED) to complete time synchronization, followed by a fixed-length competition window (Collision Window, CW); (2) After receiving the beacon, the station that has data to be sent and is in the ready state sends a signal requesting access to the gateway, that is, the time slot segment occupied by the requesting access; (3) The gateway broadcasts the occupancy of each time slot after the CW is closed; (4) After receiving the feedback frame (Feed-Back, FB) from the gateway, the site that has sent the transmission request judges the occupancy of the time slot: if it is free, it enters the data transmission queue (Data Transmission Queue, DTQ) and waits to send Data frame, otherwise enter the conflict resolution queue (Contention Resolution Queue, CRQ) and wait for the next CW to open to compete for the time slot again; (5) The station in DTQ estimates the backoff time according to its position in the queue, enters the sleep state within this time, and sends data as soon as the time is up; each data transmission is completed, followed by CW and feedback frames of the same length , the stations in the CRQ compete for the time slot again when the next CW arrives; The multi-site channel access control scheme uses time division multiplexing to divide the shared channel into four periods: downlink announcement, uplink contention, downlink confirmation and uplink data transmission; During the downlink notification period, the central gateway broadcasts the beacon and allocates the number of contention slots during the uplink contention period; The access station randomly selects a contention time slot during the uplink contention period and sends a request access signal to the gateway; During the downlink confirmation period, the gateway broadcasts the contention status of all contention slots; During the uplink data transmission period, the access station chooses to either send data immediately, or delay sending data, or contend for the channel again, or delay to contend for the channel again, according to the contention state. 2. The centralized conflict resolution method of random multiple access of a kind of wireless internet of things according to claim 1, it is characterized in that: the number of contention time slots in the downlink notification period is determined by the gateway according to the number of access stations and the uplink data request The number is determined, and the value ranges from 3 to 32. 3. A centralized conflict resolution method for random multiple access of wireless Internet of Things according to claim 1, characterized in that: the random selection of the contention time slots in the uplink contention period is independently performed by the access station in a consistent distribution Pseudo-random number calculation, the value range is distributed between 1 and the maximum number of time slots. 4. A centralized conflict resolution method for wireless Internet of Things random multiple access according to claim 1, characterized in that: the contention state of the downlink confirmation period is determined by the gateway according to the received request from the access site The access signal is determined as one of success, collision and idle. 5. The centralized conflict resolution method of random multiple access of a kind of wireless internet of things according to claim 1, it is characterized in that: during the uplink data transmission period, the access station selects to send data, then enters DTQ; The condition is that the state of the contention time slot selected by the station is successful and it is at the head of the DTQ team; the condition of delaying sending data is that the state of the contention time slot selected by the station is successful but not at the head of the DTQ team. 6. The centralized conflict resolution method of a random multiple access of wireless internet of things according to claim 1, characterized in that: during the uplink data transmission period, the access station chooses to contend for the channel, then enters CRQ; contends again The condition of the channel is that the state of the contention time slot selected by the station is conflict and it is at the head of the CRQ team; the condition of delaying the channel again is that the state of the contention time slot selected by the station is conflict but not at the head of the CRQ team.